期刊论文

Identification of the Different Degree of Moldy Indica Rice Based on Hyperspectral Technology

Wen, T.

中文

中国粮油学报

1003-0174

2017

32

8

123-130

10.3969/j.issn.1003-0174.2017.08.020

31401281:国家自然科学基金 14JJ3115:湖南省自然科学基金 2014207:湖南省高等学校科技创新团队支持计划 201410538014:国家级大学生创新创业训练计划项目

本校为第一机构

为解决快速、无损鉴别籼稻霉变程度问题。该文利用高光谱技术采集正常、轻度霉变、中度霉变和重度霉变共4类籼稻样本的光谱数据,经不同预处理后,通过连续投影算法(SPA)提取特征波长,采用多元线性回归判别分析(MLR-DA)建立籼稻霉变程度鉴别模型。同时,通过竞争性自适应重加权算法(CARS)筛选特征变量,采用偏最小二乘法回归判别分析(PLS-DA)建立鉴别模型。研究结果表明,SG-SPA-MLR-DA和RAW-CARS-PLS-DA模型的验证集相关系数均大于0.950。其中,RAW-CARS-PLS-DA模型的预测效果最优,其验证集相关系数R_P为0.969,均方根误差RMSEP为0.269,对未知籼稻样本的总体分类准确率为93.33%。该模型对不同霉变程度...

In order to solve the problem of fast and nondestructive identification of moldy, the spectral data of four kinds of moldy indica rice, including normal indica rice, mildly moldy indica rice, moderately moldy indica rice and seriously moldy indica rice, was collected by using the hyperspectral technology. After preprocessing with five methods, the successive projections algorithm (SPA) was proposed to extract the characteristic wavelength; then, the identification models of different degree of moldy indica rice were established by using the mul...